Abstract:In order to accurately identify the main interference variables and the internal coupling relations in the control process of rubber compound extrusion machine, to better implement temperature and pressure coupled system accurate control for extruder, RBF neural network was used to make system identification research. At the same time, combining with PSO algorithm, introducing coding, hybridization, crossover and mutation concepts in GA algorithm, a hybrid PSO algorithm was designed to optimize RBF neural network, the precision online identification for temperature and pressure coupling system was completed. The neural network training was carried out with the help of MATLAB software to identify the coupling relationship of the system.Meanwhile, the results identified by the neural network weight optimization were compared with those of hybrid PSO algorithm. The experimental results show that the training effect of RBF neural network is better when the hybrid PSO algorithm is adopted to optimize the RBF neural network, and the system identification of RBF neural network can be realized with high precision. The hybrid PSO algorithm optimized RBF neural network is applied to the extruder temperature and pressure control system identification, which can improve the identification accuracy of the system and the intelligence level of the extruder to a certain extent.